Commit Graph

8297 Commits

Author SHA1 Message Date
Dehao Chen 77079003dd Add indirect call promotion to SamplePGO
Summary: This patch adds metadata for indirect call promotion in the sample profile loader.

Reviewers: xur, davidxl, dnovillo

Reviewed By: davidxl

Subscribers: llvm-commits

Differential Revision: https://reviews.llvm.org/D28923

llvm-svn: 292672
2017-01-20 22:56:07 +00:00
Easwaran Raman 12585b0148 Improve PGO support for the new inliner
This adds the following to the new PM based inliner in PGO mode:

* Use block frequency analysis to derive callsite's profile count and use
that to adjust thresholds of hot and cold callsites.

* Incrementally update the BFI of the caller after a callee gets inlined
into it. This incremental update is only within an invocation of the run
method - BFI is not preserved across calls to run.
Update the function entry count of the callee after inlining it into a
caller.

* I've tuned the thresholds for the hot and cold callsites using a hacked
up version of the old inliner that explicitly computes BFI on a set of
internal benchmarks and spec. Once the new PM based pipeline stabilizes
(IIRC Chandler mentioned there are known issues) I'll benchmark this
again and adjust the thresholds if required.
Inliner PGO support.

Differential revision: https://reviews.llvm.org/D28331

llvm-svn: 292666
2017-01-20 22:44:04 +00:00
Sanjay Patel 0c1c70aef4 [ValueTracking] recognize variations of 'clamp' to improve codegen (PR31693)
By enhancing value tracking, we allow an existing min/max canonicalization to
kick in and improve codegen for several targets that have min/max instructions.

Unfortunately, recognizing min/max in value tracking may cause us to hit
a hack in InstCombiner::visitICmpInst() more often:
http://lists.llvm.org/pipermail/llvm-dev/2017-January/109340.html
...but I'm hoping we can remove that soon.

Correctness proofs based on Alive:

Name: smaxmin
Pre: C1 < C2
%cmp2 = icmp slt i8 %x, C2
%min = select i1 %cmp2, i8 %x, i8 C2
%cmp3 = icmp slt i8 %x, C1
%r = select i1 %cmp3, i8 C1, i8 %min
=>
%cmp2 = icmp slt i8 %x, C2
%min = select i1 %cmp2, i8 %x, i8 C2
%cmp1 = icmp sgt i8 %min, C1
%r = select i1 %cmp1, i8 %min, i8 C1

Name: sminmax
Pre: C1 > C2
%cmp2 = icmp sgt i8 %x, C2
%max = select i1 %cmp2, i8 %x, i8 C2
%cmp3 = icmp sgt i8 %x, C1
%r = select i1 %cmp3, i8 C1, i8 %max
=>
%cmp2 = icmp sgt i8 %x, C2
%max = select i1 %cmp2, i8 %x, i8 C2
%cmp1 = icmp slt i8 %max, C1
%r = select i1 %cmp1, i8 %max, i8 C1

---------------------------------------- 
Optimization: smaxmin 
Done: 1 
Optimization is correct! 
---------------------------------------- 
Optimization: sminmax 
Done: 1 
Optimization is correct!



Name: umaxmin
Pre: C1 u< C2
%cmp2 = icmp ult i8 %x, C2
%min = select i1 %cmp2, i8 %x, i8 C2
%cmp3 = icmp ult i8 %x, C1
%r = select i1 %cmp3, i8 C1, i8 %min
=>
%cmp2 = icmp ult i8 %x, C2
%min = select i1 %cmp2, i8 %x, i8 C2
%cmp1 = icmp ugt i8 %min, C1
%r = select i1 %cmp1, i8 %min, i8 C1

Name: uminmax
Pre: C1 u> C2
%cmp2 = icmp ugt i8 %x, C2
%max = select i1 %cmp2, i8 %x, i8 C2
%cmp3 = icmp ugt i8 %x, C1
%r = select i1 %cmp3, i8 C1, i8 %max
=>
%cmp2 = icmp ugt i8 %x, C2
%max = select i1 %cmp2, i8 %x, i8 C2
%cmp1 = icmp ult i8 %max, C1
%r = select i1 %cmp1, i8 %max, i8 C1

---------------------------------------- 
Optimization: umaxmin 
Done: 1 
Optimization is correct! 
---------------------------------------- 
Optimization: uminmax 
Done: 1 
Optimization is correct! 
 

llvm-svn: 292660
2017-01-20 22:18:47 +00:00
Sanjay Patel 5eb113d2af [InstCombine] add tests to show missed canonicalization of min/max; NFC
Unfortunately, recognizing these in value tracking may cause us to hit
a hack in InstCombiner::visitICmpInst() more often:
http://lists.llvm.org/pipermail/llvm-dev/2017-January/109340.html

...but besides being the obviously Right Thing To Do, there's a clear 
codegen win from identifying these patterns for several targets.

llvm-svn: 292655
2017-01-20 21:49:41 +00:00
Peter Collingbourne f04a390099 LowerTypeTests: Implement importing of type identifiers.
To import a type identifier we read the summary and create external
references to the symbols defined when exporting.

Differential Revision: https://reviews.llvm.org/D28546

llvm-svn: 292654
2017-01-20 21:49:34 +00:00
Daniel Berlin 68af279533 NewGVN: Remove pr31686.ll, it is tested by pr31594.ll, which is much smaller and simpler
llvm-svn: 292649
2017-01-20 21:04:58 +00:00
Daniel Berlin 26addef1a0 NewGVN: Fix PR 31686 and PR 31698 by rewriting store leader handling.
Summary:

This rewrites store expression/leader handling.  We no longer use the
value operand as the leader, instead, we store it separately.  We also
now store the stored value as part of the expression, and compare it
when comparing stores for equality.  This enables us to get rid of a
bunch of our previous hacks and machinations, as the existing
machinery takes care of everything *except* updating the stored value
on classes.  The only time we have to update it is if the storecount
goes to 0, and when we do, we destroy it.

Since we no longer use the value operand as the leader, during elimination, we have to use the value operand.  Doing this also fixes a bunch of store forwarding cases we were missing.

Any value operand we use is guaranteed to either be updated by previous eliminations, or minimized by future ones.

(IE the fact that we don't use the most dominating value operand when it's not a constant does not affect anything).

Sadly, this change also exposes that we didn't pay attention to the
output of the pr31594.ll test, as it also very clearly exposes the
same store leader bug we are fixing here.

(I added pr31682.ll anyway, but maybe we think that's too large to be useful)

On the plus side, propagate-ir-flags.ll now passes due to the
corrected store forwarding.

This change was 3 stage'd on darwin and linux, with the full test-suite.

Reviewers:
davide
Subscribers:
llvm-commits

llvm-svn: 292648
2017-01-20 21:04:30 +00:00
Haicheng Wu 201b191b82 Recommit "[InlineCost] Use TTI to check if GEP is free." #3
This is the third attemp to recommit r292526.

The original summary:

Currently, a GEP is considered free only if its indices are all constant.
TTI::getGEPCost() can give target-specific more accurate analysis. TTI is
already used for the cost of many other instructions.

llvm-svn: 292633
2017-01-20 18:51:22 +00:00
Alexey Bataev 4fe77b9329 [SLP] Initial test for fix of PR31690.
llvm-svn: 292631
2017-01-20 18:40:21 +00:00
Simon Pilgrim a50a93fcd0 [InstCombine][X86] Add MULDQ/MULUDQ undef handling
llvm-svn: 292627
2017-01-20 18:20:30 +00:00
Alexey Bataev f5677329a6 [SLP] A new test for horizontal vectorization for non-power-of-2
instructions.

llvm-svn: 292626
2017-01-20 18:04:29 +00:00
Simon Pilgrim 06f125230f [InstCombine][SSE] Tests showing missed opportunities to handle muldq/muludq with undef arguments
Fixed a typo in existing test names at the same time

llvm-svn: 292619
2017-01-20 17:06:38 +00:00
Haicheng Wu 71ef5bc0ff Revert "Recommit "[InlineCost] Use TTI to check if GEP is free." #2"
This reverts commit r292616 because the test case still has problem.

llvm-svn: 292618
2017-01-20 16:52:22 +00:00
Haicheng Wu 8f34ae2aae Recommit "[InlineCost] Use TTI to check if GEP is free." #2
This is the second attemp to recommit r292526.

The original summary:

Currently, a GEP is considered free only if its indices are all constant.
TTI::getGEPCost() can give target-specific more accurate analysis. TTI is
already used for the cost of many other instructions.

llvm-svn: 292616
2017-01-20 16:36:34 +00:00
Simon Pilgrim 2817b476e8 [InstCombine][SSE] Tests showing missed opportunities to constant fold packss/packus
llvm-svn: 292609
2017-01-20 13:21:30 +00:00
Simon Pilgrim 8942722cbc [InstCombine][SSE] Tests showing missed opportunities to handle packss/packus with undef arguments
llvm-svn: 292601
2017-01-20 11:28:07 +00:00
Simon Pilgrim 51b3b98e3a [InstCombine][SSE] Add DemandedElts support for PACKSS/PACKUS instructions
Simplify a packss/packus truncation based on the elements of the mask that are actually demanded.

Differential Revision: https://reviews.llvm.org/D28777

llvm-svn: 292591
2017-01-20 09:28:21 +00:00
Chandler Carruth e9b18e3d34 [PM] Port LoopSink to the new pass manager.
Like several other loop passes (the vectorizer, etc) this pass doesn't
really fit the model of a loop pass. The critical distinction is that it
isn't intended to be pipelined together with other loop passes. I plan
to add some documentation to the loop pass manager to make this more
clear on that side.

LoopSink is also different because it doesn't really need a lot of the
infrastructure of our loop passes. For example, if there aren't loop
invariant instructions causing a preheader to exist, there is no need to
form a preheader. It also doesn't need LCSSA because this pass is
only involved in sinking invariant instructions from a preheader into
the loop, not reasoning about live-outs.

This allows some nice simplifications to the pass in the new PM where we
can directly walk the loops once without restructuring them.

Differential Revision: https://reviews.llvm.org/D28921

llvm-svn: 292589
2017-01-20 08:42:19 +00:00
Daniel Berlin 89fea6fd9d NewGVN: Fix PR 31682, an overactive assert.
Part of the assert has been left active for further debugging.
The other part has been turned into a stat for tracking for the
moment.

llvm-svn: 292583
2017-01-20 06:38:41 +00:00
Mohammad Shahid 5dc021bf45 [SLP] Add a base test for jumbled store
Change-Id: I905ce08a02c76a6896dcfd9629547417c99adc4a
llvm-svn: 292581
2017-01-20 06:05:33 +00:00
Haicheng Wu 8f2aca388b Revert "Recommit "[InlineCost] Use TTI to check if GEP is free.""
This reverts commit r292570.  The test still has problem.

llvm-svn: 292572
2017-01-20 03:40:41 +00:00
Haicheng Wu 1af1f071ea Recommit "[InlineCost] Use TTI to check if GEP is free."
This recommits r292526 which is reverted in r292529 after fixing the test case.

The original summary:

Currently, a GEP is considered free only if its indices are all constant.
TTI::getGEPCost() can give target-specific more accurate analysis. TTI is
already used for the cost of many other instructions.

llvm-svn: 292570
2017-01-20 03:09:11 +00:00
Anna Thomas 698f0deea9 [AliasAnalysis] Fences do not modify constant memory location
Summary:
Fence instructions are currently marked as `ModRef` for all memory locations.

We can improve this for constant memory locations (such as constant globals),
since fence instructions cannot modify these locations.

This helps us to forward constant loads across fences (added test case in GVN).
There were no changes in behaviour for similar test cases in early-cse and licm.

Reviewers: dberlin, sanjoy, reames

Subscribers: llvm-commits

Differential Revision: https://reviews.llvm.org/D28914

llvm-svn: 292546
2017-01-20 00:21:33 +00:00
Davide Italiano 6c2c3e07bf [SCCP] Teach the pass how to handle `div` with overdefined operands.
This can prove that:

extern int f;
int g() {
    int x = 0;
    for (int i = 0; i < 365; ++i) {
        x /= f;
    }
    return x;
}

always returns zero. Thanks to Sanjoy for confirming this
transformation actually made sense (bugs are mine).

llvm-svn: 292531
2017-01-19 23:07:51 +00:00
Haicheng Wu e036df4723 Revert "[InlineCost] Use TTI to check if GEP is free."
This reverts commit r292526.  The test case has problem.

llvm-svn: 292529
2017-01-19 22:51:03 +00:00
Haicheng Wu da556345dc [InlineCost] Use TTI to check if GEP is free.
Currently, a GEP is considered free only if its indices are all constant.
TTI::getGEPCost() can give target-specific more accurate analysis. TTI is
already used for the cost of many other instructions.

Differential Revision: https://reviews.llvm.org/D28693

llvm-svn: 292526
2017-01-19 22:28:34 +00:00
Xin Tong 5ee40ba400 Improve what can be promoted in LICM.
Summary:
In case of non-alloca pointers, we check for whether it is a pointer
from malloc-like calls and it is not captured. In such case, we can
promote the pointer, as the caller will have no way to access this pointer
even if there is unwinding in middle of the loop.

Reviewers: hfinkel, sanjoy, reames, eli.friedman

Subscribers: llvm-commits

Differential Revision: https://reviews.llvm.org/D28834

llvm-svn: 292510
2017-01-19 19:31:40 +00:00
Davide Italiano 2ef8c4e708 [InstCombine] Simplify gep (gep p, a), (b-a)
Patch by Andrea Canciani.

Differential Revision:  https://reviews.llvm.org/D27413

llvm-svn: 292506
2017-01-19 18:51:56 +00:00
Sanjay Patel 291c3d8ff2 [InstCombine] icmp Pred (shl nsw X, C1), C0 --> icmp Pred X, C0 >> C1
Try harder to fold icmp with shl nsw as discussed here:
http://lists.llvm.org/pipermail/llvm-dev/2017-January/108749.html

This is similar to the 'shl nuw' transforms that were added with D25913.

This may eventually help solve:
https://llvm.org/bugs/show_bug.cgi?id=30773

Differential Revision: https://reviews.llvm.org/D28406

llvm-svn: 292492
2017-01-19 16:12:10 +00:00
Peter Collingbourne 22d9d3cdce LowerTypeTests: Implement exporting of type identifiers.
Type identifiers are exported by:
- Adding coarse-grained information about how to test the type
  identifier to the summary.
- Creating symbols in the object file (aliases and absolute symbols)
  containing fine-grained information about the type identifier.

Differential Revision: https://reviews.llvm.org/D28424

llvm-svn: 292462
2017-01-19 01:20:11 +00:00
Michael Kuperstein 230867e583 [LV] Run loop-simplify and LCSSA explicitly instead of "requiring" them
This changes the vectorizer to explicitly use the loopsimplify and lcssa utils,
instead of "requiring" the transformations as if they were analyses.

This is not NFC, since it changes the LCSSA behavior - we no longer run LCSSA
for all loops, but rather only for the loops we expect to modify.

Differential Revision: https://reviews.llvm.org/D28868

llvm-svn: 292456
2017-01-19 00:42:28 +00:00
Sanjay Patel cfb8a45942 [InstCombine] add tests for shl nsw with icmp eq/ne; NFCI
These should be fixed with D28406.

llvm-svn: 292441
2017-01-18 21:31:21 +00:00
Peter Collingbourne 20a00933fb ThinLTOBitcodeWriter: Clear comdats on filtered globals.
Differential Revision: https://reviews.llvm.org/D28839

llvm-svn: 292431
2017-01-18 20:03:02 +00:00
Michael Kuperstein 7cefb409b0 [LV] Allow reductions that have several uses outside the loop
We currently check whether a reduction has a single outside user. We don't
really need to require that - we just need to make sure a single value is
used externally. The number of external users of that value shouldn't actually
matter.

Differential Revision: https://reviews.llvm.org/D28830

llvm-svn: 292424
2017-01-18 19:02:52 +00:00
Alexey Bataev f86cca1a42 [SLP] Add a tests for a fix for PR30787.
Add a test for PR30787: Failure to beneficially vectorize 'copyable'
elements in integer binary ops.

llvm-svn: 292416
2017-01-18 18:07:46 +00:00
Simon Pilgrim fe2c0ed4cf [InstCombine][AVX2] Add DemandedElts support for VPERMD/VPERMPS shuffles
Simplify a vpermv shuffle mask based on the elements of the mask that are actually demanded.

llvm-svn: 292371
2017-01-18 14:47:49 +00:00
Simon Pilgrim 970d67c653 [InstCombine][AVX2] Tests showing missed opportunities to pass demanded elts through a vpermd/vpermps shuffle
llvm-svn: 292368
2017-01-18 14:23:06 +00:00
Chandler Carruth d50c5fb13f [PM] Teach LoopDeletion to correctly update the LPM when loops are
deleted.

I've expanded its test coverage a bit including adding one test that
will crash clearly without this change.

llvm-svn: 292332
2017-01-18 02:41:26 +00:00
Chandler Carruth 88c36d7852 [LoopDeletion] (cleanup, NFC) Make this test actually test what it
claims to test.

LoopSimplify was unifying the multiple exits in this test case, making
it never even test the multiple exit handling of LoopDeletion. Doh.

Now it works (thanks to a great idea from mkuper) and will fail if we
ever change something to make it stop working.

llvm-svn: 292331
2017-01-18 02:29:35 +00:00
Xin Tong 58e8142f0e 2 returns next to each other =). NFC
llvm-svn: 292315
2017-01-18 00:26:17 +00:00
Sanjay Patel 0e9f681dee [InstCombine] add tests to show missed shrinkage; NFC
A patch to partially solve this:
https://reviews.llvm.org/D28625

llvm-svn: 292296
2017-01-18 00:03:23 +00:00
Dehao Chen c3f87f02b1 Introduce -unroll-partial-threshold to separate PartialThreshold from Threshold in loop unorller.
Summary: Partial unrolling should have separate threshold with full unrolling.

Reviewers: efriedma, mzolotukhin

Reviewed By: efriedma, mzolotukhin

Subscribers: llvm-commits

Differential Revision: https://reviews.llvm.org/D28831

llvm-svn: 292293
2017-01-17 23:39:33 +00:00
Matthew Simpson e2c9ad9483 [LV] Add requires asserts to test case
llvm-svn: 292280
2017-01-17 22:21:33 +00:00
Xin Tong 0bc2977874 Add a test case for LICM when promoting locals that may be read after the throw within the loop. NFCI.
Summary: Add a test case for LICM when promoting locals that may be read after the throw within the loop.

Reviewers: eli.friedman, hfinkel, sanjoy

Subscribers: llvm-commits

Differential Revision: https://reviews.llvm.org/D28822

llvm-svn: 292261
2017-01-17 21:26:36 +00:00
Matthew Simpson 3fbdaa5906 [LV] Mark non-consecutive-like pointers non-uniform
If a memory instruction will be vectorized, but it's pointer operand is
non-consecutive-like, the instruction is a gather or scatter operation. Its
pointer operand will be non-uniform. This should fix PR31671.

Reference: https://llvm.org/bugs/show_bug.cgi?id=31671
Differential Revision: https://reviews.llvm.org/D28819

llvm-svn: 292254
2017-01-17 20:51:39 +00:00
Xin Tong 8343b5096d Rename scalar_promote.ll to scalar-promote.ll and scalar_promote-unwind.ll to scalar-promote-unwind.ll. NFCI
llvm-svn: 292251
2017-01-17 20:28:36 +00:00
Sanjoy Das 6de072a712 [EarlyCSE] Don't DSE across readnone functions that may throw
Summary: Depends on D28740

Reviewers: dberlin, chandlerc, hfinkel, majnemer

Subscribers: mcrosier, llvm-commits

Differential Revision: https://reviews.llvm.org/D28741

llvm-svn: 292249
2017-01-17 20:15:47 +00:00
Sanjay Patel f42a955c3e [InstCombine] add tests for shl nsw + icmp sle; NFC
We want to handle these cases similarly to icmp sgt, so add the tests for it.
See: https://reviews.llvm.org/D28406

llvm-svn: 292248
2017-01-17 20:15:26 +00:00
Chandler Carruth b6e32daa81 [PM] Teach the LoopPassManager to automatically canonicalize loops by
runnig LCSSA over them prior to running the loop pipeline.

This also teaches the loop PM to verify that LCSSA form is preserved
throughout the pipeline's run across the loop nest.

Most of the test updates just leverage this new functionality. One has to be
relaxed with the new PM as IVUsers is less powerful when it sees LCSSA input.

Differential Revision: https://reviews.llvm.org/D28743

llvm-svn: 292241
2017-01-17 19:18:12 +00:00
Sanjay Patel 9666996563 [ValueTracking] recognize a 'not' of an assumed condition as false
Also, add the corresponding match to the AssumptionCache's 'Affected Values' list.

Differential Revision: https://reviews.llvm.org/D28485

llvm-svn: 292239
2017-01-17 18:15:49 +00:00
David Majnemer de55c606d1 [InstCombine] Fold ((C1 OP zext(X)) & C2) -> zext((C1 OP X) & C2)
This further extends r292179 to support additional binary operators
beyond subtraction.

llvm-svn: 292238
2017-01-17 18:08:06 +00:00
Chad Rosier 8520429bdd [ValueTracking] Extend known bits to understand @llvm.bitreverse.
Differential Revision: https://reviews.llvm.org/D28780

llvm-svn: 292233
2017-01-17 17:23:51 +00:00
Simon Pilgrim d4eb800b03 [InstCombine][X86][AVX] Add DemandedElts support for VPERMILPD/VPERMILPS instructions
Simplify a vpermilvar shuffle mask based on the elements of the mask that are actually demanded.

llvm-svn: 292209
2017-01-17 11:35:03 +00:00
Sanjoy Das 679bc32c6a [InstCombine] Don't DSE across readnone functions that may throw
Summary: Depends on D28740

Reviewers: dberlin, chandlerc, hfinkel, majnemer

Subscribers: mcrosier, llvm-commits

Differential Revision: https://reviews.llvm.org/D28742

llvm-svn: 292197
2017-01-17 05:45:09 +00:00
David Majnemer 36d382b773 [InstCombine] Fold ((C1-zext(X)) & C2) -> zext((C1-X) & C2)
This is valid if C2 fits within the bitwidth of X thanks to two's
complement modulo arithmetic.

llvm-svn: 292179
2017-01-17 00:45:57 +00:00
Matt Arsenault c8cc2be9f8 Add comment to test file I forgot to save
llvm-svn: 292178
2017-01-17 00:35:28 +00:00
Matt Arsenault b948b4d8df SimplifyLibCalls: Remove checks for fabs
Use the intrinsic instead of emitting the libcall which
will be replaced by the intrinsic.

llvm-svn: 292176
2017-01-17 00:30:31 +00:00
Matt Arsenault 7233344c28 SimplifyLibCalls: Replace fabs libcalls with intrinsics
Add missing fabs(fpext) optimzation that worked with the call,
and also fixes it creating a second fpext when there were multiple
uses.

llvm-svn: 292172
2017-01-17 00:10:40 +00:00
Simon Pilgrim a0b0b96d83 [InstCombine][AVX] Tests showing missed opportunities to pass demanded elts through a permilpd/permilps shuffle mask
llvm-svn: 292165
2017-01-16 21:34:22 +00:00
Sanjay Patel ab8b32de71 [InstCombine] use m_APInt to allow shift-shift folds for vectors with splat constants
Some existing 'FIXME' tests are still not folded because of splat holes in value tracking.

llvm-svn: 292151
2017-01-16 19:35:45 +00:00
Sanjay Patel cd06f6fe10 [InstCombine] add tests to show missed vector folds; NFC
The shift-shift possibilities became easier to see after:
https://reviews.llvm.org/rL292145

llvm-svn: 292150
2017-01-16 19:23:34 +00:00
Simon Pilgrim 87eddf9aaf [InstCombine][SSE] Tests showing missed opportunities to pass demanded elts through a packss/packus truncation
llvm-svn: 292144
2017-01-16 17:26:23 +00:00
Simon Pilgrim 73a68c25a0 [InstCombine][SSE] Add DemandedElts support for PSHUFB instructions
Simplify a pshufb shuffle mask based on the elements of the mask that are actually demanded.

Differential Revision: https://reviews.llvm.org/D28745

llvm-svn: 292101
2017-01-16 11:30:41 +00:00
Sanjay Patel 21347ffddf [InstCombine] add tests to show missed vector folds; NFC
Also, add comments and remove bogus comment.

llvm-svn: 292082
2017-01-15 23:45:03 +00:00
Simon Pilgrim 21c2a50237 [InstCombine][SSE] Tests showing missed opportunities to pass demanded elts through a pshufb shuffle mask
llvm-svn: 292072
2017-01-15 17:49:04 +00:00
Sanjay Patel 5f8451afad [InstCombine] use m_APInt to allow ashr folds for vectors with splat constants
llvm-svn: 292064
2017-01-15 16:38:19 +00:00
Sanjay Patel fba2df8d0f [InstCombine] add explanatory comments to tests; NFC
llvm-svn: 292063
2017-01-15 16:22:26 +00:00
Chandler Carruth 0952750fae [PM] Clean up the testing for IVUsers, especially with the new PM.
First, I've moved a test of IVUsers from the LSR tree to a dedicated
IVUsers test directory. I've also simplified its RUN line now that the
new pass manager's loop PM is providing analyses on their own.

No functionality changed, but it makes subsequent changes cleaner.

llvm-svn: 292060
2017-01-15 09:29:27 +00:00
Chandler Carruth 5edfd4d99e [PM] Fix instcombine's analysis preservation in the new pass manager to
cover domtree and alias analysis. These are the pretty clear analyses
that we would always want to survive this pass.

To make these survive, we also need to preserve the assumption cache.

Added a test that verifies the important bits of this preservation.

llvm-svn: 292037
2017-01-14 23:25:22 +00:00
Sanjay Patel f8ed0a5e93 [InstCombine] add test to show missed vector fold; NFC
llvm-svn: 292035
2017-01-14 23:12:29 +00:00
Daniel Berlin 01831b7db2 NewGVN: Fix PR31613 test regex naming
llvm-svn: 291979
2017-01-13 23:54:10 +00:00
Sanjay Patel 40f401776b [InstCombine] optimize unsigned icmp of increment
Allows LLVM to optimize sequences like the following:

%add = add nuw i32 %x, 1
%cmp = icmp ugt i32 %add, %y

Into:

%cmp = icmp uge i32 %x, %y

Previously, only signed comparisons were being handled.

Decrements could also be handled, but 'sub nuw %x, 1' is currently canonicalized to
'add %x, -1' in InstCombineAddSub, losing the nuw flag. Removing that canonicalization
seems like it might have far-reaching ramifications so I kept this simple for now.

Patch by Matti Niemenmaa!

Differential Revision: https://reviews.llvm.org/D24700

llvm-svn: 291975
2017-01-13 23:25:46 +00:00
Sanjay Patel 2d4b456427 [InstCombine] use m_APInt to allow lshr folds for vectors with splat constants
llvm-svn: 291972
2017-01-13 23:04:10 +00:00
Sanjay Patel d511dde2ec [InstCombine / InstSimplify] add and move tests for lshr transforms; NFC
llvm-svn: 291970
2017-01-13 22:54:12 +00:00
Daniel Berlin c0431fd02d NewGVN: Move leaders around properly to ensure we have a canonical dominating leader. Fixes PR 31613.
Summary:
This is a testcase where phi node cycling happens, and because we do
not order the leaders by domination or anything similar, the leader
keeps changing.

Using std::set for the members is too expensive, and we actually don't
need them sorted all the time, only at leader changes.

We could keep both a set and a vector, and keep them mostly sorted and
resort as necessary, or use a set and a fibheap, but all of this seems
premature.

After running some statistics, we are able to avoid the vast majority
of sorting by keeping a "next leader" field.  Most congruence classes only have
leader changes once or twice during GVN.

Reviewers: davide

Subscribers: llvm-commits

Differential Revision: https://reviews.llvm.org/D28594

llvm-svn: 291968
2017-01-13 22:40:01 +00:00
David Majnemer bba17390c7 [LoopStrengthReduce] Don't bother rewriting PHIs in catchswitch blocks
The catchswitch instruction cannot be split, don't bother trying to
rewrite it.

This fixes PR31627.

llvm-svn: 291966
2017-01-13 22:24:27 +00:00
Sanjay Patel b22f6c5f26 [InstCombine] use m_APInt to allow shl folds for vectors with splat constants
llvm-svn: 291934
2017-01-13 18:39:09 +00:00
Sanjay Patel bbc1c1e46b [InstCombine] add tests to show missing transforms for vector shl; NFC
llvm-svn: 291926
2017-01-13 18:27:23 +00:00
Sanjay Patel 5178363687 [InstCombine] if the condition of a select may be known via assumes, eliminate the select
This is a limited solution for PR31512:
https://llvm.org/bugs/show_bug.cgi?id=31512

The motivation is that we will need to increase usage of llvm.assume and/or metadata to solve PR28430:
https://llvm.org/bugs/show_bug.cgi?id=28430

...and this kind of simplification is needed to take advantage of that extra information.

The 'not' test case would be handled by:
https://reviews.llvm.org/D28485

Differential Revision:
https://reviews.llvm.org/D28337

llvm-svn: 291915
2017-01-13 17:02:42 +00:00
Adam Nemet 6117caab58 Move test of lazy BFI with ORE to a generic directory
llvm-svn: 291862
2017-01-13 00:16:23 +00:00
Piotr Padlewski 9530883e8c [Devirtualization] MemDep returns non-local !invariant.group dependencies
Summary:
Memory Dependence Analysis was limited to return only local dependencies
for invariant.group handling. Now it returns NonLocal when it finds it
and then by asking getNonLocalPointerDependency we get found dep.

Thanks to this we are able to devirtualize loops!

    void indirect(A &a, int n) {
      for (int i = 0 ; i < n; i++)
        a.foo();

    }
    void test(int n) {
      A a;
      indirect(a);
    }

After inlining a.foo() will be changed to direct call, even if foo and A::A()
is external (but only if vtable definition is be available).

Reviewers: nlewycky, dberlin, chandlerc, rsmith

Subscribers: mehdi_amini, davide, llvm-commits

Differential Revision: https://reviews.llvm.org/D28137

llvm-svn: 291762
2017-01-12 11:33:58 +00:00
Chandler Carruth b4d9a310d2 Make a test actually test what it set out to test.
This test seems to have largely been relying on asserts being tripped.
It had a very specific and somewhat uninteresting grep of the output,
but it never really did anything to cause SCEV to be preserved across
loop simplify, certainly not explicitly. And a later addition to it
actually added CHECK lines despite the test never running FileCheck.

Now we actually print SCEV before and after loop simplify to make sure
it is *changing* and being *updated*. Which seems to be much more likely
the point of the test.

llvm-svn: 291740
2017-01-12 03:49:07 +00:00
Michael Kuperstein 991c2e0e57 Add test that verifies we don't peel loops in optsize functions. NFC.
llvm-svn: 291708
2017-01-11 21:42:51 +00:00
Peter Collingbourne 7636532c1b LowerTypeTests: Represent the memory region size with the constant size-1.
This means that we can use a shorter instruction sequence in the case where
the size is a power of two and on the boundary between two representations.

Differential Revision: https://reviews.llvm.org/D28421

llvm-svn: 291706
2017-01-11 21:32:10 +00:00
Peter Collingbourne 6bca5a0d82 Re-apply r291205, "LowerTypeTests: Split the pass in two: a resolution phase and a lowering phase.", with a fix for an off-by-one error.
llvm-svn: 291699
2017-01-11 20:28:46 +00:00
Daniel Berlin f6eba4be2c NewGVN: Fix PR31594, by tracking the store count of congruence
classes, and updating checking to allow for equivalence through
reachability.

(Sadly, the checking here is not perfect, and can't be made perfect,
so we'll have to disable it after we are satisfied with correctness.
Right now it is just "very unlikely" to happen.)

llvm-svn: 291698
2017-01-11 20:22:36 +00:00
Rong Xu 20f5df1d70 Resubmit "[PGO] Turn off comdat renaming in IR PGO by default"
This patch resubmits the changes in r291588.

llvm-svn: 291696
2017-01-11 20:19:41 +00:00
Michael Kuperstein f69e64662b [SLP] Remove bogus assert.
The removed assert seems bogus - it's perfectly legal for the roots of the
vectorized subtrees to be equal even if the original scalar values aren't,
if the original scalars happen to be equivalent.

This fixes PR31599.

Differential Revision: https://reviews.llvm.org/D28539

llvm-svn: 291692
2017-01-11 19:23:57 +00:00
Ivan Krasin 42e6b4fd98 Revert rL291205 because it breaks Chrome tests under CFI.
Summary:
Revert LowerTypeTests: Split the pass in two: a resolution phase and a lowering phase.

This change separates how type identifiers are resolved from how intrinsic
calls are lowered. All information required to lower an intrinsic call
is stored in a new TypeIdLowering data structure. The idea is that this
data structure can either be initialized using the module itself during
regular LTO, or using the module summary in ThinLTO backends.

Original URL: https://reviews.llvm.org/D28341

Reviewers: pcc

Subscribers: mehdi_amini, llvm-commits

Differential Revision: https://reviews.llvm.org/D28532

llvm-svn: 291684
2017-01-11 16:54:04 +00:00
Mohammed Agabaria 81d0f17055 [X86] fixing failed test in commit: r291657
Missing Requires asserts.

llvm-svn: 291659
2017-01-11 09:03:11 +00:00
Mohammed Agabaria 2c96c43388 [X86] updating TTI costs for arithmetic instructions on X86\SLM arch.
updated instructions:
pmulld, pmullw, pmulhw, mulsd, mulps, mulpd, divss, divps, divsd, divpd, addpd and subpd.

special optimization case which replaces pmulld with pmullw\pmulhw\pshuf seq. 
In case if the real operands bitwidth <= 16.

Differential Revision: https://reviews.llvm.org/D28104 

llvm-svn: 291657
2017-01-11 08:23:37 +00:00
Adam Nemet e2aaf3a35e [LICM] Report failing to hoist conditionally-executed loads
These are interesting again because the user may not be aware that this
is a common reason preventing LICM.

A const is removed from an instruction pointer declaration in order to
pass it to ORE.

Differential Revision: https://reviews.llvm.org/D27940

llvm-svn: 291649
2017-01-11 04:39:49 +00:00
Adam Nemet 81941b3195 [LICM] Report failing to hoist a load with an invariant address
These are interesting because lack of precision in alias information
could be standing in the way of this optimization.

An example is the case in the test suite that I showed in the DevMeeting
talk:

http://lab.llvm.org:8080/artifacts/opt-view_test-suite/build/MultiSource/Benchmarks/FreeBench/distray/CMakeFiles/distray.dir/html/_org_test-suite_MultiSource_Benchmarks_FreeBench_distray_distray.c.html#L236

canSinkOrHoistInst is also used from LoopSink, which does not use
opt-remarks so we need to take ORE as an optional argument.

Differential Revision: https://reviews.llvm.org/D27939

llvm-svn: 291648
2017-01-11 04:39:45 +00:00
Adam Nemet 358433ce1b [LICM] Report successful hoist/sink/promotion
Differential Revision: https://reviews.llvm.org/D27938

llvm-svn: 291646
2017-01-11 04:39:35 +00:00
Matt Arsenault 1e0edbf03c InstSimplify: Eliminate fabs on known positive
llvm-svn: 291624
2017-01-11 00:33:24 +00:00
Rong Xu acd6360251 Revert "[PGO] Turn off comdat renaming in IR PGO by default"
This patch reverts r291588: [PGO] Turn off comdat renaming in IR PGO by default,
as we are seeing some hash mismatches in our internal tests.

llvm-svn: 291621
2017-01-10 23:54:31 +00:00
Matt Arsenault fdb78f8bae InstCombine: fdiv -x, -y -> fdiv x, y
llvm-svn: 291611
2017-01-10 23:08:54 +00:00
Michael Kuperstein ee31cbe35f [LV] Don't panic when encountering the IV of an outer loop.
Bail out instead of asserting when we encounter this situation,
which can actually happen.

The reason the test uses the new PM is that the "bad" phi, incidentally, gets
cleaned up by LoopSimplify. But LICM can create this kind of phi and preserve
loop simplify form, so the cleanup has no chance to run.

This fixes PR31190.
We may want to solve this in a less conservative manner, since this phi is
actually uniform within the inner loop (or we may want LICM to output a cleaner
promotion to begin with).

Differential Revision: https://reviews.llvm.org/D28490

llvm-svn: 291589
2017-01-10 19:32:30 +00:00
Rong Xu ef1adad938 [PGO] Turn off comdat renaming in IR PGO by default
Summary:
In IR PGO we append the function hash to comdat functions to avoid the
potential hash mismatch. This turns out not legal in some cases: if the comdat
function is address-taken and used in comparison. Renaming changes the semantic.

This patch turns off comdat renaming by default.

To alleviate the hash mismatch issue, we now rename the profile variable
for comdat functions. Profile allows co-existing multiple versions of profiles
with different hash value. The inlined copy will always has the correct profile
counter. The out-of-line copy might not have the correct count. But we will
not have the bogus mismatch warning.

Reviewers: davidxl

Subscribers: llvm-commits, xur

Differential Revision: https://reviews.llvm.org/D28416

llvm-svn: 291588
2017-01-10 19:30:20 +00:00
Davide Italiano f8711f093e [SimplifyLibCalls] Propagate fast math flags while optimizing pow().
llvm-svn: 291577
2017-01-10 18:02:05 +00:00
Serge Pavlov 0668cd2c95 [StructurizeCfg] Update dominator info.
In some cases StructurizeCfg updates root node, but dominator info
remains unchanges, it causes crash when expensive checks are enabled.
To cope with this problem a new method was added to DominatorTreeBase
that allows adding new root nodes, it is called in StructurizeCfg to
put dominator tree in sync.

This change fixes PR27488.

Differential Revision: https://reviews.llvm.org/D28114

llvm-svn: 291530
2017-01-10 02:50:47 +00:00
Davide Italiano 472684eaf5 [SimplifyLibCalls] pow(x, -0.5) -> 1.0 / sqrt(x).
Differential Revision:  https://reviews.llvm.org/D28479

llvm-svn: 291486
2017-01-09 21:55:23 +00:00
Sanjay Patel 8f4910e26a [InstCombine] add test to show missed fold using llvm.assume; NFC
llvm-svn: 291472
2017-01-09 20:18:30 +00:00
Sanjay Patel eaa143c98c [InstCombine] regenerate checks; NFC
llvm-svn: 291469
2017-01-09 19:43:26 +00:00
Sanjay Patel 87495eb8ef [InstCombine] regenerate checks; NFC
llvm-svn: 291464
2017-01-09 19:18:46 +00:00
Sanjay Patel ced8fdd42a [InstCombine] remove unnecessary attribute comments from test files; NFC
llvm-svn: 291463
2017-01-09 19:13:38 +00:00
Matthew Simpson cf796478e9 [LV] Fix-up external IV users after updating dominator tree
This patch delays the fix-up step for external induction variable users until
after the dominator tree has been properly updated. This should fix PR30742.
The SCEVExpander in InductionDescriptor::transform can generate code in the
wrong location if the dominator tree is not up-to-date. We should work towards
keeping the dominator tree up-to-date throughout the transformation.

Reference: https://llvm.org/bugs/show_bug.cgi?id=30742
Differential Revision: https://reviews.llvm.org/D28168

llvm-svn: 291462
2017-01-09 19:05:29 +00:00
Xin Tong c13a8e84d1 Intrinsic::Bitreverse is safe to speculate
Summary: Intrinsic::Bitreverse is safe to speculate

Reviewers: hfinkel, mkuper, arsenm, jmolloy

Subscribers: llvm-commits

Differential Revision: https://reviews.llvm.org/D28471

llvm-svn: 291456
2017-01-09 17:57:08 +00:00
Daniel Berlin b755aea8eb NewGVN: Fix PR 31573, a failure to verify memory congruency due to
not excluding ourselves when checking if any equivalent stores
exist.

llvm-svn: 291421
2017-01-09 05:34:29 +00:00
Piotr Padlewski 09ad678bc4 [MemDep] NFC walk invariant.group graph only down
Summary:
By using stripPointerCasts we can get to the root
value and then walk down the bitcast graph

Reviewers: reames

Subscribers: llvm-commits

Differential Revision: https://reviews.llvm.org/D28181

llvm-svn: 291405
2017-01-08 22:26:06 +00:00
Matt Arsenault a7d2194168 SimplifyLibCalls: Remove incorrect optimization of fabs
fabs(x * x) is not generally safe to assume x is positive if x is a NaN.
This is also less general than it could be, so this will be replaced
with a transformation on the intrinsic.

llvm-svn: 291359
2017-01-07 19:55:12 +00:00
Daniel Berlin 32f8d560dd NewGVN: Make sure we properly lookup operand leaders while creating
congruence classes for stores, and then keep them up to date.  Add
testcases.

llvm-svn: 291351
2017-01-07 16:55:14 +00:00
Daniel Berlin d92e7f9f74 NewGVN: Fix PR 31501.
Summary: LLVM's non-standard notion of phi nodes means we can't both try to substitute for undef in phi nodes *and* use phi nodes as leaders all the time. This changes NewGVN to use the same semantics as SimplifyPHINode to decide which phi nodes are equivalent.

Reviewers: davide

Subscribers: llvm-commits

Differential Revision: https://reviews.llvm.org/D28312

llvm-svn: 291308
2017-01-07 00:01:42 +00:00
David Majnemer 63da0c238b [InstSimplify] Optimize away udivs in the presence of range metadata
We know that udiv %V, C can be optimized away to 0 if %V is ult C.

llvm-svn: 291296
2017-01-06 22:58:02 +00:00
David Majnemer 8c0e62f507 [InstSimplify] Optimize away urems in the presence of range metadata
We know that urem %V, C can be optimized away to %V if %V is ult C.

llvm-svn: 291282
2017-01-06 21:23:51 +00:00
Mehdi Amini 27d224fbbb Fix LoopLoadElimination to keep original alignment on the inital hoisted store
This is fixing a bug where Loop Vectorization is widening a load but
with a lower alignment. Hoisting the load without propagating the alignment
will allow inst-combine to later deduce a higher alignment that what the pointer
actually is.

Differential Revision: https://reviews.llvm.org/D28408

llvm-svn: 291281
2017-01-06 21:06:51 +00:00
Sanjay Patel 2715d92389 [InstCombine] add a vector version of a test added in r291262; NFC
llvm-svn: 291265
2017-01-06 19:14:05 +00:00
Sanjay Patel 8d4aa10960 [InstCombine] move and add tests for icmp + shl nsw; NFC
As discussed here:
http://lists.llvm.org/pipermail/llvm-dev/2017-January/108749.html
...we should be able to better optimize this pattern.

llvm-svn: 291262
2017-01-06 18:57:54 +00:00
Peter Collingbourne 81271b7bd2 LowerTypeTests: Split the pass in two: a resolution phase and a lowering phase.
This change separates how type identifiers are resolved from how intrinsic
calls are lowered. All information required to lower an intrinsic call
is stored in a new TypeIdLowering data structure. The idea is that this
data structure can either be initialized using the module itself during
regular LTO, or using the module summary in ThinLTO backends.

Differential Revision: https://reviews.llvm.org/D28341

llvm-svn: 291205
2017-01-06 02:22:47 +00:00
Michael Kuperstein c9acad12e9 [LICM] Allow promotion of some stores that are not guaranteed to execute.
Promotion is always legal when a store within the loop is guaranteed to execute.

However, this is not a necessary condition - for promotion to be memory model
semantics-preserving, it is enough to have a store that dominates every exit
block. This is because if the store dominates every exit block, the fact the
exit block was executed implies the original store was executed as well.

Differential Revision: https://reviews.llvm.org/D28147

llvm-svn: 291171
2017-01-05 20:42:06 +00:00
Mohammed Agabaria 23599ba794 Currently isLikelyComplexAddressComputation tries to figure out if the given stride seems to be 'complex' and need some extra cost for address computation handling.
This code seems to be target dependent which may not be the same for all targets.
Passed the decision whether the given stride is complex or not to the target by sending stride information via SCEV to getAddressComputationCost instead of 'IsComplex'.

Specifically at X86 targets we dont see any significant address computation cost in case of the strided access in general.

Differential Revision: https://reviews.llvm.org/D27518

llvm-svn: 291106
2017-01-05 14:03:41 +00:00
Peter Collingbourne b2ce2b6805 IR: Module summary representation for type identifiers; summary test scaffolding for lowertypetests.
Set up basic YAML I/O support for module summaries, plumb the summary into
the pass and add a few command line flags to test YAML I/O support. Bitcode
support to come separately, as will the code in LowerTypeTests that actually
uses the summary. Also add a couple of tests that pass by virtue of the pass
doing nothing with the summary (which happens to be the correct thing to do
for those tests).

Differential Revision: https://reviews.llvm.org/D28041

llvm-svn: 291069
2017-01-05 03:39:00 +00:00
Sanjay Patel 95faecb766 [InstSimplify] add tests to show missing select simplifications; NFC
llvm-svn: 291043
2017-01-05 00:40:52 +00:00
Wolfgang Pieb ce13e716c5 [DWARF] Null out the debug locs of load instructions that have been moved by GVN
performing partial redundancy elimination (PRE). Not doing so can cause jumpy line
tables and confusing (though correct) source attributions.

Differential Revision: https://reviews.llvm.org/D27857

llvm-svn: 291037
2017-01-04 23:58:26 +00:00
Matt Arsenault 3bdd75d01e InstCombine: Fold cos(-x) -> cos(x)
Also cos(fabs(x)) -> cos(x)

llvm-svn: 291022
2017-01-04 22:49:03 +00:00
Michael Kuperstein f381f35977 Add positive test for sqrt "partial inlining". NFC.
llvm-svn: 291009
2017-01-04 21:24:56 +00:00
Michael Kuperstein 020af9c258 Remove accidentally target-dependent test and pacify bots.
llvm-svn: 291004
2017-01-04 21:08:53 +00:00
Michael Kuperstein fc74da13a9 Add positive test for sqrt "partial inlining". NFC.
llvm-svn: 291001
2017-01-04 20:48:30 +00:00
Simon Pilgrim 6cfb5caf05 Revert r290970 [SLPVectorizer] Regenerate test.
The check script will use var names before they are declared, which filecheck doesn't like.

llvm-svn: 290971
2017-01-04 16:12:07 +00:00
Simon Pilgrim 4629b46bba [SLPVectorizer] Regenerate test.
Missed var name

llvm-svn: 290970
2017-01-04 16:01:55 +00:00
Simon Pilgrim 1d5b0377af Regenerate test.
llvm-svn: 290969
2017-01-04 15:52:41 +00:00
Chandler Carruth 96809ae7ea [Inliner] Fix a test where I typo'ed 'CHECK' as 'CHCEK' when converting
to FileCheck.

Fortunately, it passes. =]

Spotted in review by Bob Wilson!

llvm-svn: 290953
2017-01-04 11:15:01 +00:00
David Majnemer b5e365c970 [InstCombine] Add a test for r290733
llvm-svn: 290929
2017-01-04 02:21:37 +00:00
David Majnemer cb892e9066 [InstCombine] Move casts around shift operations
It is possible to perform a left shift before zero extending if the
shift would only shift out zeros.

llvm-svn: 290928
2017-01-04 02:21:34 +00:00
David Majnemer 022d2a563b [InstCombine] Combine adds across a zext
We can perform the following:
(add (zext (add nuw X, C1)), C2) -> (zext (add nuw X, C1+C2))

This is only possible if C2 is negative and C2 is greater than or equal to negative C1.

llvm-svn: 290927
2017-01-04 02:21:31 +00:00
Matt Arsenault 56ff4839ae InstCombine: Fold fabs on select of constants
llvm-svn: 290913
2017-01-03 22:40:34 +00:00
Sanjay Patel ada846aff0 [InstCombine] tighten checks for tests of assume -> metadata transform; NFC
llvm-svn: 290903
2017-01-03 19:32:11 +00:00
Xin Tong 883dd1b6c4 Enable disabled loopidiom test. Apparently we handle it now
Summary:
Enable disabled loopidiom test. Apparently we handle it now.
Maybe due to improvements to AA.

Reviewers: atrick, danielcdh, hfinkel

Subscribers: llvm-commits

Differential Revision: https://reviews.llvm.org/D28171

llvm-svn: 290900
2017-01-03 19:08:05 +00:00
Matt Arsenault b264c94963 InstCombine: Add fma with constant transforms
DAGCombine already does these.

llvm-svn: 290860
2017-01-03 04:32:35 +00:00
Matt Arsenault 1cc294c85d InstCombine: Add fma + fabs/fneg transforms
fma (fneg x), (fneg y), z -> fma x, y, z
fma (fabs x), (fabs x), z -> fma x, x, z

llvm-svn: 290859
2017-01-03 04:32:31 +00:00
Xin Tong 2940231ff0 Make sure total loop body weight is preserved in loop peeling
Summary:
Regardless how the loop body weight is distributed, we should preserve
total loop body weight. i.e. we should have same weight reaching the body of the loop
or its duplicates in peeled and unpeeled case.

Reviewers: mkuper, davidxl, anemet

Subscribers: llvm-commits

Differential Revision: https://reviews.llvm.org/D28179

llvm-svn: 290833
2017-01-02 20:27:23 +00:00
Daniel Berlin aa0ec1e992 NewGVN: Add a test case for equivalent phis.
llvm-svn: 290830
2017-01-02 19:55:13 +00:00
Daniel Berlin 43a5f998df NewGVN: Add forgotten testcase for PR 31483
llvm-svn: 290829
2017-01-02 19:49:20 +00:00
Sanjay Patel 0e3ae439cf [InstCombine] add explanatory comment to test; NFC
The test was added at r290797, and a patch to enable the transform is proposed in D28204.

llvm-svn: 290798
2017-01-01 18:20:49 +00:00
Sanjay Patel 07537c2b6e [InstCombine] add test to show potential nonnull attribute propagation; NFC
This will change with the current draft of:
https://reviews.llvm.org/D28204

llvm-svn: 290797
2017-01-01 17:18:00 +00:00
Craig Topper d00db69227 [InstCombine][AVX-512] Teach InstCombine that llvm.x86.avx512.vcomi.sd and llvm.x86.avx512.vcomi.ss don't use the upper elements of their input.
This was already done for the SSE/SSE2 version of the intrinsics.

llvm-svn: 290776
2016-12-31 00:45:06 +00:00
Piotr Padlewski da36215017 [MemDep] Handle gep with zeros for invariant.group
Summary:
gep 0, 0 is equivalent to bitcast. LLVM canonicalizes it
to getelementptr because it make SROA can then handle it.

Simple case like

    void g(A &a) {
        z(a);
        if (glob)
            a.foo();
    }
    void testG() {
        A a;
        g(a);
    }

was not devirtualized with -fstrict-vtable-pointers because luck of
handling for gep 0 in Memory Dependence Analysis

Reviewers: dberlin, nlewycky, chandlerc

Subscribers: llvm-commits

Differential Revision: https://reviews.llvm.org/D28126

llvm-svn: 290763
2016-12-30 18:45:07 +00:00
Michael Kuperstein 76e06c8858 [LICM] When promoting scalars, allow inserting stores to thread-local allocas.
This is similar to the allocfn case - if an alloca is not captured, then it's
necessarily thread-local.

Differential Revision: https://reviews.llvm.org/D28170

llvm-svn: 290738
2016-12-30 01:03:17 +00:00
Dehao Chen cc76344ef5 Use continuous boosting factor for complete unroll.
Summary:
The current loop complete unroll algorithm checks if unrolling complete will reduce the runtime by a certain percentage. If yes, it will apply a fixed boosting factor to the threshold (by discounting cost). The problem for this approach is that the threshold abruptly. This patch makes the boosting factor a function of runtime reduction percentage, capped by a fixed threshold. In this way, the threshold changes continuously.

The patch also simplified the code by reducing one parameter in UP.

The patch only affects code-gen of two speccpu2006 benchmark:

445.gobmk binary size decreases 0.08%, no performance change.
464.h264ref binary size increases 0.24%, no performance change.

Reviewers: mzolotukhin, chandlerc

Subscribers: llvm-commits

Differential Revision: https://reviews.llvm.org/D26989

llvm-svn: 290737
2016-12-30 00:50:28 +00:00
Daniel Berlin e0bd37e78f NewGVN: Fix PR 31491 by ensuring that we touch the right instructions. Change to one based numbering so we can assert we don't cause the same bug again.
llvm-svn: 290724
2016-12-29 22:15:12 +00:00
Craig Topper b57a84dace [InstCombine] Fix some of the AVX-512 scalar arithmetic test cases to do a better job of testing what they intended to test.
The accidentally had trivially dead code. Also needed to adjust the rounding mode to not CUR_DIRECTION so the intrinsics don't get converted to native operations before going through SimplifyDemandedVectorElts.

llvm-svn: 290702
2016-12-29 02:29:04 +00:00
Chandler Carruth 05ca5acc9e [PM] Introduce a devirtualization iteration layer for the new PM.
This is an orthogonal and separated layer instead of being embedded
inside the pass manager. While it adds a small amount of complexity, it
is fairly minimal and the composability and control seems worth the
cost.

The logic for this ends up being nicely isolated and targeted. It should
be easy to experiment with different iteration strategies wrapped around
the CGSCC bottom-up walk using this kind of facility.

The mechanism used to track devirtualization is the simplest one I came
up with. I think it handles most of the cases the existing iteration
machinery handles, but I haven't done a *very* in depth analysis. It
does however match the basic intended semantics, and we can tweak or
tune its exact behavior incrementally as necessary. One thing that we
may want to revisit is freshly building the value handle set on each
iteration. While I don't think this will be a significant cost (it is
strictly fewer value handles but more churn of value handes than the old
call graph), it is conceivable that we'll want a somewhat more clever
tracking mechanism. My hope is to layer that on as a follow up patch
with data supporting any implementation complexity it adds.

This code also provides for a basic count heuristic: if the number of
indirect calls decreases and the number of direct calls increases for
a given function in the SCC, we assume devirtualization is responsible.
This matches the heuristics currently used in the legacy pass manager.

Differential Revision: https://reviews.llvm.org/D23114

llvm-svn: 290665
2016-12-28 11:07:33 +00:00
Chandler Carruth 443e57e01d [PM] Teach the CGSCC's CG update utility to more carefully invalidate
analyses when we're about to break apart an SCC.

We can't wait until after breaking apart the SCC to invalidate things:
1) Which SCC do we then invalidate? All of them?
2) Even if we invalidate all of them, a newly created SCC may not have
   a proxy that will convey the invalidation to functions!

Previously we only invalidated one of the SCCs and too late. This led to
stale analyses remaining in the cache. And because the caching strategy
actually works, they would get used and chaos would ensue.

Doing invalidation early is somewhat pessimizing though if we *know*
that the SCC structure won't change. So it turns out that the design to
make the mutation API force the caller to know the *kind* of mutation in
advance was indeed 100% correct and we didn't do enough of it. So this
change also splits two cases of switching a call edge to a ref edge into
two separate APIs so that callers can clearly test for this and take the
easy path without invalidating when appropriate. This is particularly
important in this case as we expect most inlines to be between functions
in separate SCCs and so the common case is that we don't have to so
aggressively invalidate analyses.

The LCG API change in turn needed some basic cleanups and better testing
in its unittest. No interesting functionality changed there other than
more coverage of the returned sequence of SCCs.

While this seems like an obvious improvement over the current state, I'd
like to revisit the core concept of invalidating within the CG-update
layer at all. I'm wondering if we would be better served forcing the
callers to handle the invalidation beforehand in the cases that they
can handle it. An interesting example is when we want to teach the
inliner to *update and preserve* analyses. But we can cross that bridge
when we get there.

With this patch, the new pass manager an build all of the LLVM test
suite at -O3 and everything passes. =D I haven't bootstrapped yet and
I'm sure there are still plenty of bugs, but this gives a nice baseline
so I'm going to increasingly focus on fleshing out the missing
functionality, especially the bits that are just turned off right now in
order to let us establish this baseline.

llvm-svn: 290664
2016-12-28 10:34:50 +00:00
Chandler Carruth 9900d18bab [PM] Teach the inliner's call graph update to handle inserting new edges
when they are call edges at the leaf but may (transitively) be reached
via ref edges.

It turns out there is a simple rule: insert everything as a ref edge
which is a safe conservative default. Then we let the existing update
logic handle promoting some of those to call edges.

Note that it would be fairly cheap to make these call edges right away
if that is desirable by testing whether there is some existing call path
from the source to the target. It just seemed like slightly more
complexity in this code path that isn't strictly necessary. If anyone
feels strongly about handling this differently I'm happy to change it.

llvm-svn: 290649
2016-12-28 03:13:12 +00:00
Michael Kuperstein cd7ad7130f [InstCombine] Canonicalize insert splat sequences into an insert + shuffle
This adds a combine that canonicalizes a chain of inserts which broadcasts
a value into a single insert + a splat shufflevector.

This fixes PR31286.

Differential Revision: https://reviews.llvm.org/D27992

llvm-svn: 290641
2016-12-28 00:18:08 +00:00
Bryant Wong 7cb744621b [MemCpyOpt] Don't sink LoadInst below possible clobber.
Differential Revision: https://reviews.llvm.org/D26811

llvm-svn: 290611
2016-12-27 17:58:12 +00:00
Chandler Carruth 625038d5d5 [PM] Turn on the new PM's inliner in addition to the current one for
most of the inliner test cases.

The inliner involves a bunch of interesting code and tends to be where
most of the issues I've seen experimenting with the new PM lie. All of
these test cases pass, but I'd like to keep some more thorough coverage
here so doing a fairly blanket enabling.

There are a handful of interesting tests I've not enabled yet because
they're focused on the always inliner, or on functionality that doesn't
(yet) exist in the inliner.

llvm-svn: 290592
2016-12-27 07:18:43 +00:00
Chandler Carruth 141bf5d14d [PM] Add one of the features left out of the initial inliner patch:
skipping indirectly recursive inline chains.

To do this, we implicitly build an inline stack for each callsite and
check prior to inlining that doing so would not form a cycle. This uses
the exact same technique and even shares some code with the legacy PM
inliner.

This solution remains deeply unsatisfying to me because it means we
cannot actually iterate the inliner externally. Doing so would not be
able to easily detect and avoid such cycles. Some day I would very much
like to have a solution that works without this internal state to detect
cycles, but this is not that day.

llvm-svn: 290590
2016-12-27 06:46:20 +00:00
Chandler Carruth db6ced8484 [PM] Wire up another test to the new pass manager.
Nothing really interesting here, but I had to improve the test to use
variables rather than hard coding value names as we happen to end up
with different value names in the new PM.

llvm-svn: 290589
2016-12-27 06:46:16 +00:00
George Burgess IV ed16024a9b [Analysis] Ignore `nobuiltin` on `allocsize` function calls.
We currently ignore the `allocsize` attribute on functions calls with
the `nobuiltin` attribute when trying to lower `@llvm.objectsize`. We
shouldn't care about `nobuiltin` here: `allocsize` is explicitly added
by the user, not inferred based on a function's symbol.

llvm-svn: 290588
2016-12-27 06:32:14 +00:00
Craig Topper 72f2d4e8d6 [InstCombine][X86] Add DemandedElts support for 512-bit PMULDQ/PMULUDQ instructions
PMULDQ/PMULUDQ vXi64 instructions only use the even numbered v2Xi32 input elements which SimplifyDemandedVectorElts should try and use.

This builds on r290554 which added supported for 128 and 256-bit.

llvm-svn: 290582
2016-12-27 05:30:09 +00:00
Chandler Carruth 03130d981c [PM] Teach the inliner in the new PM to merge attributes after inlining.
Also enable the new PM in the attributes test case which caught this
issue.

llvm-svn: 290572
2016-12-27 03:39:54 +00:00
Chandler Carruth 62c8b81ea8 [Inliner] Modernize all of the inliner tests that were using grep.
This mostly involved converting from grep to FileCheck and tidying up
the IR used.

In one case (invoke_test-3.ll) the test had become completely pointless
as we use 'resume' rather than 'unwind' now, and even then it did not
occur at the end of the line.

llvm-svn: 290570
2016-12-27 02:47:37 +00:00
Craig Topper 7f8540b5e7 [AVX-512][InstCombine] Teach InstCombine to turn masked scalar add/sub/mul/div with rounding intrinsics into normal IR operations if the rounding mode is CUR_DIRECTION.
An earlier commit added support for unmasked scalar operations. At that time isel wouldn't generate an optimal sequence for masked operations, but that has now been fixed.

llvm-svn: 290566
2016-12-27 01:56:30 +00:00
Craig Topper a0439377e6 [InstCombine][AVX-512] Add masked scalar add/sub/mul/div intrinsic test cases that don't have a CUR_DIRECTION rounding mode.
The CUR_DIRECTION case will be optimized in a future commit so this provides coverage for the other cases.

llvm-svn: 290565
2016-12-27 01:56:27 +00:00
Chandler Carruth 0ee8bb11c3 [PM] Move the collection of call sites to a more appropriate place
inside of `InlineFunction`. Prior to this, call instructions are
specifically being rewritten and replaced within the inlined region,
invalidating some of the call sites.

Several of these regions are using the same technique to walk the
inlined region so this seems clearly safe up to this point.

I've also added a short circuit to the scan for call sites based on what
other code is doing.

With this, the most common crash I've found in the new inliner code is
fixed. I've turned it on for another test case that covers this
scenario.

I'll make my way through most of the other inliner test cases
just to get some easy coverage next.

llvm-svn: 290562
2016-12-27 01:24:50 +00:00
Craig Topper 020b228155 [AVX-512][InstCombine] Teach InstCombine to turn packed add/sub/mul/div with rounding intrinsics into normal IR operations if the rounding mode is CUR_DIRECTION.
llvm-svn: 290559
2016-12-27 00:23:16 +00:00
Chandler Carruth 6e9bb7e064 [PM] Teach the always inliner in the new pass manager to support
removing fully-dead comdats without removing dead entries in comdats
with live members.

This factors the core logic out of the current inliner's internals to
a reusable utility and leverages that in both places. The factored out
code should also be (minorly) more efficient in cases where we have very
few dead functions or dead comdats to consider.

I've added a test case to cover this behavior of the always inliner.
This is the last significant bug in the new PM's always inliner I've
found (so far).

llvm-svn: 290557
2016-12-26 23:43:27 +00:00
Simon Pilgrim c9cf7fc7a4 [InstCombine][X86] Add DemandedElts support for PMULDQ/PMULUDQ instructions
PMULDQ/PMULUDQ vXi64 instructions only use the even numbered v2Xi32 input elements which SimplifyDemandedVectorElts should try and use.

Differential Revision: https://reviews.llvm.org/D28119

llvm-svn: 290554
2016-12-26 23:28:17 +00:00
Daniel Berlin d59e8010c5 Don't use our own incorrect version of isTriviallyDeadInstruction in NewGVN. Fixes PR/31472
llvm-svn: 290549
2016-12-26 18:44:36 +00:00
Davide Italiano 8ea5e4fcae [NewGVN] Change test to reflect difference between GVN and NewGVN.
The current GVN algorithm folds unconditional branches to, it claims,
expose more PRE oportunities. The folding, if really needed,
(which is not sure, as it's not really proved it improves analysis)
can be done by an earlier cleanup pass instead of GVN itself.
Ack'ed/SGTM'd by Daniel Berlin.

Differential Revision:  https://reviews.llvm.org/D28117

llvm-svn: 290546
2016-12-26 18:10:09 +00:00
Bryant Wong b5e03b61e2 [InstCombiner] Simplify lib calls to `round{,f}`
Differential Revision: https://reviews.llvm.org/D28110

llvm-svn: 290542
2016-12-26 14:29:29 +00:00
Chandler Carruth 80db76d556 Test the different scenarios of GlobalDCE and comdats more
systematically and document in the test what all is going on.

This replaces the PR-named test that was the only coverage for GlobalDCE
and comdats previously. I wrote this because I wasn't certain how
comdat DCE was supposed to work and wanted to step through what
GlobalDCE did to fully understand it. After talking to folks and reading
the code and really staring at things it all makes sense but it seemed
good to help write down some of this in a more explicit and fully
covering test case.

For example, it seemed like a bug that GlobalDCE didn't consider comdat
participation of ifuncs. Specifically it seemed like an accident because
testing didn't really cover that case. But in fact, ifuncs specifically
cannot participate in a comdat despite having that API. The new test
case covers this and explicitly documents that DCE gets to fire here
even though there are comdats involved.

Also, we didn't have any positive tests for the challenging cases such
as usage cycles between comdat participants that might make them seem
alive except that there is no external edge into the cycle.

llvm-svn: 290537
2016-12-26 08:54:01 +00:00
Craig Topper 7b788ada2d [AVX-512][InstCombine] Teach InstCombine to turn scalar add/sub/mul/div with rounding intrinsics into normal IR operations if the rounding mode is CUR_DIRECTION.
Summary:
I only do this for unmasked cases for now because isel is failing to fold the mask. I'll try to fix that soon.

I'll do the same thing for packed add/sub/mul/div in a future patch.

Reviewers: delena, RKSimon, zvi, craig.topper

Subscribers: llvm-commits

Differential Revision: https://reviews.llvm.org/D27879

llvm-svn: 290535
2016-12-26 06:33:19 +00:00
Craig Topper e328045711 [AVX-512][InstCombine] Teach InstCombine to converted masked vpermv intrinsics into shufflevector instructions
Summary:
This patch adds support for converting the masked vpermv intrinsics into shufflevector instructions if the indices are constants.

We also need to wrap a select instruction around the shuffle to take care of the masking part. InstCombine will take care of optimizing the select if the mask is constant so I didn't bother checking for that.

Reviewers: zvi, delena, spatel, RKSimon

Subscribers: llvm-commits

Differential Revision: https://reviews.llvm.org/D27825

llvm-svn: 290530
2016-12-25 23:58:57 +00:00
Bryant Wong a07d9b1460 [AliasAnalysis] Teach BasicAA about memcpy.
Differential Revision: https://reviews.llvm.org/D27034

llvm-svn: 290526
2016-12-25 22:42:27 +00:00
Daniel Berlin d7c12ee54c Value number stores and memory states so we can detect when memory states are equivalent (IE store of same value to memory).
Reviewers: davide

Subscribers: llvm-commits

Differential Revision: https://reviews.llvm.org/D28084

llvm-svn: 290525
2016-12-25 22:23:49 +00:00
Simon Pilgrim 3265d951b6 [InstCombine][X86] Add tests showing missed opportunities to simplify PMULUDQ/PMULDQ inputs.
PMULUDQ/PMULDQ - only the even elements (0, 2, 4, 6) of the vXi32 inputs are required.

llvm-svn: 290502
2016-12-24 17:30:19 +00:00
Chandler Carruth cdfdd4330a [PM] Remove a bunch of junk that snuck in when I failed at manipulating
my editor to close and commit the patch. Sorry for the noise.

llvm-svn: 290460
2016-12-23 23:39:31 +00:00
Chandler Carruth 4eaff12ba2 [PM] Teach the always inlining test case to be much more strict about
whether functions are removed, and fix the new PM's always inliner to
actually pass this test.

Without this, the new PM's always inliner leaves all the functions
kicking around which won't work out very well given the semantics of
always inline.

Doing this really highlights how frustrating the current alwaysinline
semantic contract is though -- why can we put it on *external*
functions, etc?

Also I've added a number of tricky and interesting test cases for
removing functions with the always inliner. There is one remaining case
not handled -- fully removing comdats -- and I've left a FIXME about
this.

llvm-svn: 290457
2016-12-23 23:33:35 +00:00
Chandler Carruth f32f63f222 [PM] Clean up test case and comments a bit. NFC.
llvm-svn: 290456
2016-12-23 23:33:32 +00:00
Davide Italiano 34f94384a5 [LICM] Work around LICM needs to maintain state across loops.
The pass creates some state which expects to be cleaned up by
a later instance of the same pass. opt-bisect happens to expose
this not ideal design because calling skipLoop() will result in
this state not being cleaned up at times and an assertion firing
in `doFinalization()`. Chandler tells me the new pass manager will
give us options to avoid these design traps, but until it's not ready,
we need a workaround for the current pass infrastructure. Fix provided
by Andy Kaylor, see the review for a complete discussion.

Differential Revision:  https://reviews.llvm.org/D25848

llvm-svn: 290427
2016-12-23 13:12:50 +00:00
Chandler Carruth eb119ece4a Fix some DOS-style line endings that I suspect snuck in from one of the
frustrating Subversion clients that fails to do line ending translation
of text files.

llvm-svn: 290404
2016-12-23 02:02:26 +00:00
Evgeniy Stepanov 27d4c9b71b [cfi] Emit jump tables as a function-level inline asm.
Use a dummy private function with inline asm calls instead of module
level asm blocks for CFI jumptables.

The main advantage is that now jumptable codegen can be affected by
the function attributes (like target_cpu on ARM). Module level asm
gets the default subtarget based on the target triple, which is often
not good enough.

This change also uses asm constraints/arguments to reference
jumptable targets and aliases directly. We no longer do asm name
mangling in an IR pass.

Differential Revision: https://reviews.llvm.org/D28012

llvm-svn: 290384
2016-12-22 22:22:35 +00:00
Davide Italiano e05e3306a3 [NewGVN] Add the pass to PassRegistry.def.
We need to hook up here to get it working with the new PM.
Add a test while here (and remove a typo).

llvm-svn: 290350
2016-12-22 16:35:02 +00:00
Davide Italiano 7e274e02ae [GVN] Initial check-in of a new global value numbering algorithm.
The code have been developed by Daniel Berlin over the years, and
the new implementation goal is that of addressing shortcomings of
the current GVN infrastructure, i.e. long compile time for large
testcases, lack of phi predication, no load/store value numbering
etc...

The current code just implements the "core" GVN algorithm, although
other pieces (load coercion, phi handling, predicate system) are
already implemented in a branch out of tree. Once the core is stable,
we'll start adding pieces on top of the base framework.
The test currently living in test/Transform/NewGVN are a copy
of the ones in GVN, with proper `XFAIL` (missing features in NewGVN).
A flag will be added in a future commit to enable NewGVN, so that
interested parties can exercise this code easily.

Differential Revision:  https://reviews.llvm.org/D26224

llvm-svn: 290346
2016-12-22 16:03:48 +00:00
Adrian Prantl 1eadba1c8c Renumber testcase metadata nodes after r290153.
This patch renumbers the metadata nodes in debug info testcases after
https://reviews.llvm.org/D26769. This is a separate patch because it
causes so much churn. This was implemented with a python script that
pipes the testcases through llvm-as - | llvm-dis - and then goes
through the original and new output side-by side to insert all
comments at a close-enough location.

Differential Revision: https://reviews.llvm.org/D27765

llvm-svn: 290292
2016-12-22 00:45:21 +00:00
Adrian Prantl ec9ebba778 Legalize metadata in legacy testcases
llvm-svn: 290288
2016-12-21 23:38:17 +00:00
Adrian Prantl 762e4b72c6 Legalize metadata in legacy testcases
llvm-svn: 290287
2016-12-21 23:36:06 +00:00
Adrian Prantl aad5df484c Legalize metadata in legacy testcases
llvm-svn: 290286
2016-12-21 23:30:35 +00:00
David Majnemer b0761a0c1b Revert "[InstCombine] New opportunities for FoldAndOfICmp and FoldXorOfICmp"
This reverts commit r289813, it caused PR31449.

llvm-svn: 290266
2016-12-21 19:21:59 +00:00
Adam Nemet 32e6a34c02 [LDist] Match behavior between invoking via optimization pipeline or opt -loop-distribute
In r267672, where the loop distribution pragma was introduced, I tried
it hard to keep the old behavior for opt: when opt is invoked
with -loop-distribute, it should distribute the loop (it's off by
default when ran via the optimization pipeline).

As MichaelZ has discovered this has the unintended consequence of
breaking a very common developer work-flow to reproduce compilations
using opt: First you print the pass pipeline of clang
with -debug-pass=Arguments and then invoking opt with the returned
arguments.

clang -debug-pass will include -loop-distribute but the pass is invoked
with default=off so nothing happens unless the loop carries the pragma.
While through opt (default=on) we will try to distribute all loops.

This changes opt's default to off as well to match clang.  The tests are
modified to explicitly enable the transformation.

llvm-svn: 290235
2016-12-21 04:07:40 +00:00
George Burgess IV 3f08914e7e [Analysis] Centralize objectsize lowering logic.
We're currently doing nearly the same thing for @llvm.objectsize in
three different places: two of them are missing checks for overflow,
and one of them could subtly break if InstCombine gets much smarter
about removing alloc sites. Seems like a good idea to not do that.

llvm-svn: 290214
2016-12-20 23:46:36 +00:00
Chandler Carruth 1d96311447 [PM] Provide an initial, minimal port of the inliner to the new pass manager.
This doesn't implement *every* feature of the existing inliner, but
tries to implement the most important ones for building a functional
optimization pipeline and beginning to sort out bugs, regressions, and
other problems.

Notable, but intentional omissions:
- No alloca merging support. Why? Because it isn't clear we want to do
  this at all. Active discussion and investigation is going on to remove
  it, so for simplicity I omitted it.
- No support for trying to iterate on "internally" devirtualized calls.
  Why? Because it adds what I suspect is inappropriate coupling for
  little or no benefit. We will have an outer iteration system that
  tracks devirtualization including that from function passes and
  iterates already. We should improve that rather than approximate it
  here.
- Optimization remarks. Why? Purely to make the patch smaller, no other
  reason at all.

The last one I'll probably work on almost immediately. But I wanted to
skip it in the initial patch to try to focus the change as much as
possible as there is already a lot of code moving around and both of
these *could* be skipped without really disrupting the core logic.

A summary of the different things happening here:

1) Adding the usual new PM class and rigging.

2) Fixing minor underlying assumptions in the inline cost analysis or
   inline logic that don't generally hold in the new PM world.

3) Adding the core pass logic which is in essence a loop over the calls
   in the nodes in the call graph. This is a bit duplicated from the old
   inliner, but only a handful of lines could realistically be shared.
   (I tried at first, and it really didn't help anything.) All told,
   this is only about 100 lines of code, and most of that is the
   mechanics of wiring up analyses from the new PM world.

4) Updating the LazyCallGraph (in the new PM) based on the *newly
   inlined* calls and references. This is very minimal because we cannot
   form cycles.

5) When inlining removes the last use of a function, eagerly nuking the
   body of the function so that any "one use remaining" inline cost
   heuristics are immediately refined, and queuing these functions to be
   completely deleted once inlining is complete and the call graph
   updated to reflect that they have become dead.

6) After all the inlining for a particular function, updating the
   LazyCallGraph and the CGSCC pass manager to reflect the
   function-local simplifications that are done immediately and
   internally by the inline utilties. These are the exact same
   fundamental set of CG updates done by arbitrary function passes.

7) Adding a bunch of test cases to specifically target CGSCC and other
   subtle aspects in the new PM world.

Many thanks to the careful review from Easwaran and Sanjoy and others!

Differential Revision: https://reviews.llvm.org/D24226

llvm-svn: 290161
2016-12-20 03:15:32 +00:00
Adrian Prantl bceaaa9643 [IR] Remove the DIExpression field from DIGlobalVariable.
This patch implements PR31013 by introducing a
DIGlobalVariableExpression that holds a pair of DIGlobalVariable and
DIExpression.

Currently, DIGlobalVariables holds a DIExpression. This is not the
best way to model this:

(1) The DIGlobalVariable should describe the source level variable,
    not how to get to its location.

(2) It makes it unsafe/hard to update the expressions when we call
    replaceExpression on the DIGLobalVariable.

(3) It makes it impossible to represent a global variable that is in
    more than one location (e.g., a variable with multiple
    DW_OP_LLVM_fragment-s).  We also moved away from attaching the
    DIExpression to DILocalVariable for the same reasons.

This reapplies r289902 with additional testcase upgrades and a change
to the Bitcode record for DIGlobalVariable, that makes upgrading the
old format unambiguous also for variables without DIExpressions.

<rdar://problem/29250149>
https://llvm.org/bugs/show_bug.cgi?id=31013
Differential Revision: https://reviews.llvm.org/D26769

llvm-svn: 290153
2016-12-20 02:09:43 +00:00
Sanjay Patel 5a443ac000 [InstCombine] use commutative matcher for pattern with commutative operators
This is a case that was missed in:
https://reviews.llvm.org/rL290067
...and it would regress if we fix operand complexity (PR28296).

llvm-svn: 290127
2016-12-19 18:35:37 +00:00
Sanjay Patel dd46b52942 [InstCombine] add folds for icmp (umin|umax X, Y), X
This is a follow-up to:
https://reviews.llvm.org/rL289855 (https://reviews.llvm.org/D27531)
https://reviews.llvm.org/rL290111

llvm-svn: 290118
2016-12-19 17:32:37 +00:00
Florian Hahn 2e03213f90 [LoopVersioning] Require loop-simplify form for loop versioning.
Summary:
Requiring loop-simplify form for loop versioning ensures that the
runtime check block always dominates the exit block.
    
This patch closes #30958 (https://llvm.org/bugs/show_bug.cgi?id=30958).

Reviewers: silviu.baranga, hfinkel, anemet, ashutosh.nema

Subscribers: ashutosh.nema, mzolotukhin, efriedma, hfinkel, llvm-commits

Differential Revision: https://reviews.llvm.org/D27469

llvm-svn: 290116
2016-12-19 17:13:37 +00:00
Sanjay Patel 8296c6c96f [InstCombine] add folds for icmp (smax X, Y), X
This is a follow-up to:
https://reviews.llvm.org/rL289855 (D27531)

llvm-svn: 290111
2016-12-19 16:28:53 +00:00
Daniel Jasper aec2fa352f Revert @llvm.assume with operator bundles (r289755-r289757)
This creates non-linear behavior in the inliner (see more details in
r289755's commit thread).

llvm-svn: 290086
2016-12-19 08:22:17 +00:00